//Do not use in parallel as the gappedItems sorting will screw up
public List <ScoredSequence> DetailedDenovoSequencing(PatternTools.MSParser.MSFull theMS)
{
MSFull myMS = PatternTools.ObjectCopier.Clone(theMS);
PrepareMSForDeNovo(myMS);
List <GraphNode> allNodes = GetNodes(myMS);
List <List <GraphNode> > allPaths = new List <List <GraphNode> >();
foreach (GraphNode g in allNodes)
{
//foreach (List<GraphNode> dp in g.DownPaths) {
// dp.Insert(0, g);
//}
allPaths.AddRange(g.DownPaths);
}
allPaths = allPaths.Distinct().ToList();
allPaths.Sort((a, b) => EvaluatePathForIntensity(b).CompareTo(EvaluatePathForIntensity(a)));
List <ScoredSequence> theResults = new List <ScoredSequence>();
foreach (List <GraphNode> p in allPaths)
{
double avgPPM = EvaluatePathForPPMError(p);
double intensity = EvaluatePathForIntensity(p);
int sequentialScore = p.Count;
string sequence = GetSequence(p);
ScoredSequence s = new ScoredSequence();
s.AvgPPMError = Math.Round(avgPPM, 1);
s.Sequence = sequence;
s.TotalIntensity = Math.Round(intensity, 0);
s.SequentialScore = sequentialScore;
theResults.Add(s);
}
int removed = theResults.RemoveAll(a => a.Sequence.Equals(""));
Console.WriteLine("Removing bad results: " + removed);
Dictionary <string, List <ScoredSequence> > cluster = (from result in theResults
group result by result.Sequence into resultGroup
select new { theRegex = resultGroup.Key, results = resultGroup }).ToDictionary(a => a.theRegex, a => a.results.ToList());
theResults.Clear();
foreach (KeyValuePair <string, List <ScoredSequence> > kvp in cluster)
{
kvp.Value.Sort((a, b) => b.TotalIntensity.CompareTo(a.TotalIntensity));
theResults.Add(kvp.Value[0]);
}
return(theResults);
}
public static SpectrumComparisonResult Do(List <TheTests> theTests, List <PredictedIon> theoretical, List <Ion> experimentalPeaks, double ppm, int cleanedPeptideSequenceLength, double relativeIntensityThreshold)
{
PatternTools.MSParser.MSFull ms = new MSFull();
ms.MSData = experimentalPeaks;
ms.TotalIonIntensity = experimentalPeaks.Sum(a => a.Intensity);
return(Do(theTests, theoretical, ms, ppm, cleanedPeptideSequenceLength, relativeIntensityThreshold));
}
public void PlotSpectrum(MSParserLight.MSLight mslight, double ppm, string peptideSequence, List <ModificationItem> myMods, bool etdMode)
{
MSFull msFull = new MSFull();
for (int i = 0; i < mslight.MZ.Count; i++)
{
msFull.MSData.Add(new Ion(mslight.MZ[i], mslight.Intensity[i], 0, mslight.ScanNumber));
}
PlotSpectrum(msFull, ppm, peptideSequence, myMods, etdMode);
}
private void PrepareMSForDeNovo(MSFull myMS)
{
double maxIntensity = myMS.MSData.Max(b => b.Intensity);
myMS.MSData.RemoveAll(a => a.Intensity < intensityPercentageCuttoff * maxIntensity);
myMS.MSData.Sort((a, b) => a.MZ.CompareTo(b.MZ));
myMS.MSData.Insert(0, new Ion(1.007825, maxIntensity, 0, 0));
foreach (double d in myMS.DechargedPrecursorsFromZLine)
{
myMS.MSData.Add(new Ion(d, 1, 0, 0));
}
}
public void PlotSpectrum(MSFull ms, double ppm, string peptideSequence, List <ModificationItem> myMods, bool etdMode)
{
msViewer1.MyMS = ms;
msViewer1.SetMS2PPM = ppm;
msViewer1.PeptideSequence = peptideSequence;
msViewer1.Modifications = myMods;
if (etdMode)
{
msViewer1.SetToCIDMode();
}
else
{
msViewer1.SetToETDMode();
}
msViewer1.SetToCIDMode();
msViewer1.Plot();
}
//Optimized for finding most intense sequence
public double FastMaxSequentialScoreCoverage(PatternTools.MSParser.MSFull theMS)
{
MSFull myMS = PatternTools.ObjectCopier.Clone(theMS);
PrepareMSForDeNovo(myMS);
List <GraphNode> allNodes = GetNodes(myMS);
List <List <GraphNode> > allPaths = new List <List <GraphNode> >();
foreach (GraphNode g in allNodes)
{
allPaths.AddRange(g.DownPaths);
}
allPaths = allPaths.Distinct().ToList();
allPaths.Sort((a, b) => EvaluatePathForIntensity(b).CompareTo(EvaluatePathForIntensity(a)));
double coverage = allPaths[0].Max(a => a.MZ) - allPaths[0].Min(a => a.MZ);
return(coverage);
}
private List <GraphNode> GetNodes(MSFull myMS)
{
List <GraphNode> allNodes = new List <GraphNode>(myMS.MSData.Count);
//Create the Node Array
for (int i = 0; i < myMS.MSData.Count; i++)
{
allNodes.Add(new GraphNode(Math.Round(myMS.MSData[i].MZ, 3), myMS.MSData[i].Intensity, i));;
}
//Include the Child Links
for (int i = 0; i < allNodes.Count; i++)
{
for (int j = i; j < allNodes.Count; j++)
{
if (j == i)
{
continue;
}
double delta = allNodes[j].MZ - allNodes[i].MZ;
if (delta > largestGap)
{
break;
}
double thisStep = allNodes[i].MZ + delta;
//List<GapItem> gp = (from g in gapItems
// orderby (PatternTools.pTools.PPM(thisStep, allNodes[i].MZ + g.GapSize))
// select g).ToList();
foreach (GapItem gp in gapItems)
{
double ppm2;
if ((ppm2 = PatternTools.pTools.PPM(allNodes[i].MZ + delta, allNodes[i].MZ + gp.GapSize)) < ppmTolerance)
{
GraphNode.GraphNodeLink gl = new GraphNode.GraphNodeLink(j, ppm2, gp, allNodes[j].Intensity);
allNodes[i].ChildLinks.Add(gl);
}
}
}
}
//Now lets sort all childlinks by error cost
foreach (GraphNode g in allNodes)
{
//g.ChildLinks.Sort((a, b) => b.Intensity.CompareTo(a.Intensity));
g.ChildLinks.Sort((a, b) => a.ErrorCostPPM.CompareTo(b.ErrorCostPPM));
g.DownPaths = new List <List <GraphNode> >();
g.DownPaths.Add(new List <GraphNode>());
}
List <double> bestIntensities = new List <double>(resultBuffer);
bestIntensities.Add(0);
for (int i = allNodes.Count - 1; i >= 0; i--)
{
GraphNode rNode = allNodes[i];
if (rNode.ChildLinks.Count == 0)
{
continue;
}
foreach (GraphNode.GraphNodeLink gnl in rNode.ChildLinks)
{
List <List <GraphNode> > pathsToTrim = new List <List <GraphNode> >();
foreach (List <GraphNode> path in allNodes[gnl.TheNode].DownPaths)
{
List <GraphNode> newPath = new List <GraphNode>()
{
allNodes[gnl.TheNode]
};
newPath.AddRange(path);
double downIntensity = EvaluatePathForIntensity(newPath);
double bestIntensitiesMin = bestIntensities.Min();
if (downIntensity > bestIntensitiesMin)
{
bestIntensities.Add(downIntensity);
rNode.DownPaths.Add(newPath);
if (bestIntensities.Count > resultBuffer)
{
bestIntensities.Remove(bestIntensitiesMin);
}
}
else
{
pathsToTrim.Add(path);
}
}
if (pathsToTrim.Count > 0)
{
//allNodes[gnl.TheNode].DownPaths = allNodes[gnl.TheNode].DownPaths.Except(pathsToTrim).ToList();
}
}
}
foreach (GraphNode g in allNodes)
{
foreach (List <GraphNode> l in g.DownPaths)
{
l.Insert(0, g);
}
//I dont understand why this does not work
//g.DownPaths.ForEach(a => a.Insert(0, g.MyIndex));
}
return(allNodes);
}
public static SpectrumComparisonResult Do(List <TheTests> theTests, List <PredictedIon> theoretical, MSFull experimentalTmp, double ppm, int cleanedPeptideSequenceLength, double relativeIntensityThreshold)
{
SpectrumComparisonResult scr = new SpectrumComparisonResult();
//Make sure we dont screw up the original spectrum
MSFull experimental = PatternTools.ObjectCopier.Clone(experimentalTmp);
double maxIntensity = experimental.MSData.Max(a => a.Intensity);
experimental.MSData.RemoveAll(a => a.Intensity / maxIntensity < relativeIntensityThreshold);
bool allTests = false;
if (theTests.Contains(TheTests.AllTests))
{
allTests = true;
}
//Just to make sure
foreach (PredictedIon p in theoretical)
{
p.Matched = false;
}
//Before all find all matched theoretical and experimental
List <Ion> matchedIons = new List <Ion>(100);
foreach (PredictedIon pi in theoretical)
{
List <Ion> theIons = experimental.MSData.FindAll(a => PatternTools.pTools.PPM(a.MZ, pi.MZ) < ppm);
if (theIons.Count > 0)
{
matchedIons.AddRange(theIons);
pi.Matched = true;
}
}
matchedIons = matchedIons.Distinct().ToList();
if (theTests.Contains(TheTests.SignalPercentage) || allTests)
{
scr.SignalPercentage = Math.Round(matchedIons.Sum(a => a.Intensity) / experimental.TotalIonIntensity, 6);
}
if (theTests.Contains(TheTests.ACount) || allTests)
{
scr.ACount = SequentialScore(theoretical, IonSeries.A);
}
if (theTests.Contains(TheTests.BCount) || allTests)
{
scr.BCount = SequentialScore(theoretical, IonSeries.B);
}
if (theTests.Contains(TheTests.CCount) || allTests)
{
scr.CCount = SequentialScore(theoretical, IonSeries.C);
}
if (theTests.Contains(TheTests.XCount) || allTests)
{
scr.XCount = SequentialScore(theoretical, IonSeries.X);
}
if (theTests.Contains(TheTests.YCount) || allTests)
{
scr.YCount = SequentialScore(theoretical, IonSeries.Y);
}
if (theTests.Contains(TheTests.ZCount) || allTests)
{
scr.ZCount = SequentialScore(theoretical, IonSeries.Z);
}
if (theTests.Contains(TheTests.ComplementaryPairsAX) || allTests)
{
scr.ComplementaryPairsAX = CountComplementaryPairs(scr.ACount, scr.XCount, cleanedPeptideSequenceLength);
}
if (theTests.Contains(TheTests.ComplementaryPairsBY) || allTests)
{
scr.ComplementaryPairsBY = CountComplementaryPairs(scr.BCount, scr.YCount, cleanedPeptideSequenceLength);
}
if (theTests.Contains(TheTests.ComplementaryPairsCZ) || allTests)
{
scr.ComplementaryPairsCZ = CountComplementaryPairs(scr.CCount, scr.ZCount, cleanedPeptideSequenceLength);
}
return(scr);
}